Agentic AI
AI systems that pursue a goal by planning, using tools and taking multi-step actions with some autonomy, rather than only answering a single prompt.
Agentic AI describes systems built to pursue a goal rather than answer one question. An agent breaks a goal into steps, chooses tools, takes actions in other software, observes the results and adjusts, with varying levels of autonomy and human oversight.
A chatbot answers. An agent does: it searches, reads files, calls APIs, edits documents, sends messages or updates a system of record, often over many steps.
What changes with agents
- Interfaces. Agents do not need your application's screens; they need its capabilities. This puts a premium on capability readiness, and on standards such as the Model Context Protocol.
- Authority. Because an agent can act, the line between what it can do and what it may do matters: capability is not authority.
- Evaluation. The harness around the model shapes performance, so evaluation has to cover the whole system.
- Time. Agents that carry goals across days raise the stakes of persistent delegation and context debt.
Agentic transformation is therefore as much about operating models, supervision and recovery as about the model.
Read more in The next AI interface may never be seen.
Related terms
Harness
The software around a model that manages context, tool use, sub-agents and the execution environment. It shapes real-world capability, so it must be evaluated with the model, not ignored.
Persistent delegation
Handing an AI agent a goal it keeps carrying over time — remembering context, noticing relevant events and acting across applications — rather than completing a single prompted task.
Capability is not authority
A design principle for AI agents: being technically able to perform an action does not mean the agent should be permitted to perform it. Delegation needs gradients of authority.
Model Context Protocol
An open standard, usually called MCP, for connecting AI models and agents to external tools and data sources in a consistent way, so they can discover and call an organization's capabilities.
Used in these essays
Stop adopting AI
AI adoption is rising faster than enterprise value because companies keep installing new intelligence inside old operating models.
OpenAI Dots is a test of whether AI can carry a goal, not just complete a task
Dots matters less as another capable assistant than as a test of persistent delegation: can AI keep carrying a goal without giving the user a new system to manage?
The wrong questions about AI right now
Many of the questions that helped us orient ourselves around generative AI are now too blunt to be useful. The harder work is no longer asking what AI is in the abstract, but specifying where it works, where it fails, what authority it should have, and what the whole system costs.
A cheaper AI model can move the cost instead of removing it
A lower model bill can hide a higher workflow bill. The useful AI TCO question is not only what got cheaper, but where the cost moved.
Make the AI vendor demo fail
A polished AI demo proves that a system can succeed under prepared conditions. A buying decision needs different evidence: what happens when the system is wrong, blocked, uncertain or halfway through an action.
The hidden metric in AI automation is supervision
As AI moves from assisting work to leading it, hours saved stop telling the whole story. The scarce resource shifts to human supervision: approvals, exceptions, context and judgment.
The next AI interface may never be seen
Agents are turning software capabilities into an interface of their own. The next enterprise design problem is deciding what should be callable, by whom, and under which boundaries.
Your chatbot is borrowing from the next interaction
AI customer service is usually measured one interaction at a time. But a failed automated interaction can change which channel a customer chooses next time. That makes future adoption part of the economics, not a separate trust metric.